IRC chat logs for #ltsp on irc.libera.chat (webchat)
Channel log from 23 March 2008 (all times are UTC)
| 00:32 | <johnny> hmmm.. ltsp-update-image is a bit too openbsd-inetd specific ..
| |
| 00:32 | i don't even have that in portage
| |
| 00:32 | only have xinetd
| |
| 00:36 | <warren> I don't use ltsp-update-image for a number of reasons
| |
| 00:36 | I will need to write something that better suits my needs
| |
| 00:38 | <johnny> a number of reasons?
| |
| 00:39 | <warren> johnny, like xinetd, but there are others
| |
| 00:40 | <johnny> please ellaborate
| |
| 00:40 | i need to decide on a direction too
| |
| 00:40 | <warren> I don't use initramfs
| |
| 00:40 | initramfs-tools
| |
| 00:40 | <johnny> i dont use initramfs-tools yet either.. it doesn't exist in portage
| |
| 00:41 | <warren> I also NEVER need to edit my pxelinux config file
| |
| 00:41 | It just isn't necessary the way I do it
| |
| 00:43 | <johnny> hmm.. can you autodetect it somehow?
| |
| 00:44 | <warren> I suppose I can add what I need and make it conditional to that script
| |
| 00:44 | but it might be better for me to write what I need from scratch, then figure out what's truly shared between the two
| |
| 00:44 | then make a common functions file
| |
| 00:44 | that's how vagrant and I are successfully fixing many parts
| |
| 00:45 | johnny, many of the scripts in k12linux directories of ltsp-trunk are only there temporarily
| |
| 00:45 | johnny, it will migrate to common code after I'm past the "meet feature parity" stage.
| |
| 00:45 | <johnny> i'm getting closer..
| |
| 00:46 | <warren> btw, you are gentoo?
| |
| 00:46 | <johnny> yes
| |
| 00:46 | <warren> just out of personal curiosity, why gentoo for you over any other distro?
| |
| 00:46 | <johnny> altho my deployment is ubuntu
| |
| 00:46 | <warren> oh
| |
| 00:46 | <johnny> i've been using gentoo for 5 years
| |
| 00:46 | it's been a wonderful experience
| |
| 00:47 | i'm friendly with many of the devs..
| |
| 00:47 | and i love the things i can do to tweak it
| |
| 00:47 | since it's so bare, i had to learn how to setup things manually that other distros setup automagically.. since they are intended for desktop usage out of the box
| |
| 00:47 | <warren> they?
| |
| 00:47 | <johnny> now.. i can fix things when those automagic things go wrong :)
| |
| 00:48 | fedora, ubuntu,suse
| |
| 00:49 | <warren> doh
| |
| 00:49 | hmmm
| |
| 00:49 | I'm *really* close to fixing the ldm xauth problem
| |
| 00:49 | I got it working with LDM_DIRECTX=true
| |
| 00:49 | but now the default isn't working
| |
| 00:49 | <johnny> as far as gentoo and ltsp is concerned..
| |
| 00:50 | i figured this was a good way to learn how things worked internally
| |
| 00:50 | <warren> ah, learning experience
| |
| 00:50 | <johnny> and pick up a few more skills
| |
| 00:50 | i was going to switch my laptop media center box thingy to ubuntu..
| |
| 00:50 | but if i setup ltsp.. i won't have to
| |
| 00:51 | altho my main laptop is ubuntu
| |
| 00:52 | i haven't found a reason to check out fedora
| |
| 00:53 | about the only valid reason i can think of to use it.. is if i were to get a job at redhat
| |
| 00:53 | i like redhat as a company in general
| |
| 00:53 | just never tried again after the evil of redhat9
| |
| 00:53 | then again.. maybe i was just too n00b
| |
| 00:54 | at the time.. gentoo had a mostly technical community.. so searching the forums always gave me good answers
| |
| 00:54 | so i stuck with it
| |
| 00:55 | aha.. xorg-server is done
| |
| 00:55 | <warren> fedora is very different from RH9 both technically and as a community
| |
| 00:55 | we have many of the best FOSS engineers in the world and most of them are expected to work with the community
| |
| 00:56 | including key people who do upstream kernel, gcc and glibc
| |
| 00:56 | <johnny> i'm sure
| |
| 00:56 | yes i know this
| |
| 00:56 | i read things ya know? :)
| |
| 00:56 | i said i liked redhat as a company :)
| |
| 00:57 | maybe i'll put foedra9 on my box
| |
| 00:57 | fedora*
| |
| 00:57 | laptop*
| |
| 00:57 | ugggh
| |
| 00:58 | slashdot1x has joined #ltsp | |
| 00:59 | <slashdot1x> how do you prevent ehci_hcd modules from loading?
| |
| 00:59 | <warren> slashdot1x, specifically that USB driver for all usb?
| |
| 00:59 | slashdot1x, or*
| |
| 00:59 | <slashdot1x> I've tried to add new blacklist entry under /etc/modprobe.d
| |
| 00:59 | * warren digs into openssh source to attempt to figure out why it is behaving the way I see... | |
| 00:59 | <slashdot1x> but seem getting ignored
| |
| 01:00 | I have problem with certain usb pendrives
| |
| 01:00 | <warren> slashdot1x, it might be happening before it is able to read the blacklist depending on your distro
| |
| 01:00 | <slashdot1x> if I remove ehci_hcd module, it works ok
| |
| 01:00 | <warren> slashdot1x, sounds like a kernel bug
| |
| 01:00 | what distro?
| |
| 01:00 | and what kernel?
| |
| 01:00 | <slashdot1x> I'm using xubuntu 7.04
| |
| 01:01 | Linux w5 2.6.20-15-386 #2 Sun Apr 15 07:34:00 UTC 2007 i686 GNU/Linux
| |
| 01:01 | <warren> sorry, I know nothing about Ubuntu
| |
| 01:02 | slashdot1x, in some cases where blacklist fails in modprobe.d, you can use the install line instead
| |
| 01:02 | slashdot1x, install ehci_hcd /bin/true
| |
| 01:02 | <slashdot1x> what I was trying to do is, avoid running ehci_hcd (USB ver2.0) from loading
| |
| 01:03 | <warren> slashdot1x, this only works however if ehci_hcd is being loaded AFTER initramfs and while /etc/modprobe.d is available
| |
| 01:03 | <slashdot1x> I'm ok with ver 1.1 speed
| |
| 01:03 | <warren> slashdot1x, if it is loading before that, then you will need a script to unload the driver manually in addition to the above install line.
| |
| 01:05 | <slashdot1x> warren: I'll try your suggestion
| |
| 01:06 | thanks
| |
| 01:07 | I'll report back the result soon
| |
| 02:30 | gvy has joined #ltsp | |
| 02:31 | gvy_ has quit IRC | |
| 02:56 | <johnny> laga, you about?
| |
| 03:07 | K_O-Gnom has joined #ltsp | |
| 03:28 | subir has quit IRC | |
| 03:34 | praveer_cool has joined #ltsp | |
| 03:42 | chup has joined #ltsp | |
| 03:44 | chupa has quit IRC | |
| 04:04 | bobby_C has joined #ltsp | |
| 04:04 | <dberkholz> man, i really hate trying to get these wrt54g routers working.
| |
| 04:04 | there's always random combinations of stuff that's just broken
| |
| 04:09 | Solv has joined #ltsp | |
| 04:10 | subir has joined #ltsp | |
| 04:10 | <johnny> dberkholz, what software are you using?
| |
| 04:11 | <Solv> I'm trying to get kiwi-ltsp working under suse 10.3, I have ltsp 4.2 running under 10.2 fine,and client works great...but under 10.3 with kiwi-ltsp it can't find pxelinux.0 from the tftp server...the logs say: in.tftpd[4563]: tftp: client does not accept options
| |
| 04:11 | and the line before that is: in.tftpd[4563]: tftp: client does not accept options
| |
| 04:11 | but the same client works fine under ltsp 4.2???
| |
| 04:12 | * johnny no use kiwi | |
| 04:12 | <johnny> some folks do tho, i'm sure they'll be about sometime
| |
| 04:12 | it's 5am. i gotta hit the sack
| |
| 04:13 | dberkholz, i hope you're around a bit tomorrow.. i need a bit of advice :)
| |
| 04:13 | <dberkholz> probably more or less
| |
| 04:14 | <Solv> ta
| |
| 04:14 | <dberkholz> johnny: dd-wrt, fwiw
| |
| 04:44 | theboss_ has joined #ltsp | |
| 04:44 | Solv has quit IRC | |
| 04:50 | basanta has joined #ltsp | |
| 05:02 | theboss_ has quit IRC | |
| 05:04 | theboss_ has joined #ltsp | |
| 05:16 | bobby_C has quit IRC | |
| 05:34 | theboss_ has quit IRC | |
| 05:35 | theboss_ has joined #ltsp | |
| 05:35 | xcasex has quit IRC | |
| 05:43 | theboss_ has quit IRC | |
| 05:52 | ikonia has joined #ltsp | |
| 05:55 | subir has quit IRC | |
| 06:21 | praveer_cool has quit IRC | |
| 06:27 | praveer_cool has joined #ltsp | |
| 06:30 | <laga> johnny: yes
| |
| 06:36 | mikkel has joined #ltsp | |
| 06:44 | basanta has quit IRC | |
| 08:04 | daya has joined #ltsp | |
| 08:23 | chupa has joined #ltsp | |
| 08:23 | chup has quit IRC | |
| 09:02 | chup has joined #ltsp | |
| 09:04 | chupa has quit IRC | |
| 09:22 | Q-FUNK has joined #ltsp | |
| 09:22 | vagrantc has joined #ltsp | |
| 09:35 | <vagrantc> warren: regarding DISPLAYNUM being TTYNUM - 1 ...
| |
| 09:35 | warren: "it is the way it has always been and thou shalt not question it"
| |
| 09:35 | that's about the only reason i can think of.
| |
| 09:36 | i've always thought it was silly too.
| |
| 09:36 | daya has quit IRC | |
| 09:41 | <Q-FUNK> hm
| |
| 10:35 | <johnny> laga, and now? :)
| |
| 10:39 | <laga> johnny: yes
| |
| 10:40 | <johnny> you were talking about aufs right?
| |
| 10:40 | <laga> yes
| |
| 10:41 | <johnny> well.. aufs is installable via kernel module, but unionfs can't anymore
| |
| 10:41 | so i'd like to know a bit more about how you have it working
| |
| 10:44 | <laga> are you trying to use aufs and unionfs simultaneously?
| |
| 10:45 | <johnny> no
| |
| 10:46 | is there a reason i should?
| |
| 10:46 | <laga> no
| |
| 10:46 | i was just wondering where you're actually encountering problems
| |
| 10:46 | <johnny> i just told you :)
| |
| 10:46 | unionfs doesn't come with mainline, and must be patched it
| |
| 10:47 | genkernel (our initramfs creator) removed unionfs support, because it is only available via patch
| |
| 10:47 | and it was often not compiling
| |
| 10:47 | <laga> ah
| |
| 10:47 | now i get it
| |
| 10:47 | <johnny> it used to be available as a module
| |
| 10:48 | <laga> you're forced to use aufs because unionfs isn't available anymore.. and you want to know how i use it
| |
| 10:48 | <johnny> well if aufs is trouble.. we'll provide the patched kernel up front..
| |
| 10:48 | but i'd prefer not to
| |
| 10:49 | hmm.. i also had a question about your previous locales patch
| |
| 10:49 | why did you default to en_GB for locales? i saw that (before knowing who did it) and wrapped the locale-gen call
| |
| 10:50 | instead of calling it at all if the current locale is C
| |
| 10:50 | <laga> johnny: something (i dont remember what it was) failed when LANG wasn't set correctly.. so i set it to C, but ogra told me that i should change it to en_GB because perl complains otherwise
| |
| 10:50 | and because it's the default locale for ubuntu boxes
| |
| 10:51 | <johnny> oh
| |
| 10:51 | i didn't know that
| |
| 10:51 | ok..so my patch is fine for me.. and yours for you :)
| |
| 10:51 | <laga> so yeah, my first approach was setting the locale the C ;)
| |
| 10:52 | <johnny> locale-gen will fail ifyou have C
| |
| 10:52 | so i skipped the call if LANG is unset
| |
| 10:53 | <laga> i'll look for the trouble ticket
| |
| 10:53 | <johnny> well if en_GB is default for ubuntu, dont bother
| |
| 10:54 | <laga> heh
| |
| 10:54 | so what do you want to know about aufs? how i call it?
| |
| 10:55 | <johnny> yes, where you set it up, how stable it is, etc
| |
| 10:57 | Faithful has quit IRC | |
| 10:57 | <laga> i set it up in the initramfs.. it's proven to be more stable than unionfs. my copy-on-write branch is on NFS which was pretty much impossible with the unionfs (1.x) found in ubuntu
| |
| 10:58 | http://www.pastebin.ca/953994 <- here's how i mount the darn thing
| |
| 10:59 | there is a special 'real-readonly' mode in aufs which might be useful for squashfs, but i havent tried it yet
| |
| 11:00 | <johnny> i wonder if union mounts will ever make it into the kernel
| |
| 11:01 | <laga> i'd prefer aufs ;))
| |
| 11:01 | <johnny> union mounts will be at the vfs level
| |
| 11:01 | instead of on top
| |
| 11:01 | seems like al viro won't accept anything less into mainline
| |
| 11:02 | <laga> too bad :/
| |
| 11:02 | <johnny> thus the reason i want something that can be installed as a module
| |
| 11:03 | <laga> aufs requires some kernel patches in my case to work well with NFS
| |
| 11:03 | just some missing symbols need to be exported
| |
| 11:05 | Q-FUNK has quit IRC | |
| 11:05 | <johnny> uggh..
| |
| 11:05 | cyberorg has quit IRC | |
| 11:05 | <laga> :)
| |
| 11:06 | not sure if they're necessary when building a 'normal' setup..
| |
| 11:06 | cyberorg has joined #ltsp | |
| 11:07 | <johnny> normal?
| |
| 11:10 | how is yours abnormal?
| |
| 11:10 | <laga> johnny: a setup which doesnt involve FNS ;)
| |
| 11:10 | err, NFS
| |
| 11:10 | what is used for the gentoo disks?
| |
| 11:11 | <johnny> for?
| |
| 11:11 | i know it uses squashfs somehow.. and has an option for you to mount the livecd over nfs
| |
| 11:11 | <laga> for union things? or how do they do that?
| |
| 11:11 | <johnny> unionfs was just recently pulled
| |
| 11:11 | nothing has replaced it
| |
| 11:11 | <laga> ah
| |
| 11:12 | Q-FUNK has joined #ltsp | |
| 11:23 | <johnny> what kernel are you currently using ?
| |
| 11:36 | <warren> vagrantc, ok, then I'm changing it
| |
| 11:38 | vagrantc, I have changes for xauth that need to go in
| |
| 11:38 | vagrantc, I'd like to put them in after you checkin your changes
| |
| 11:39 | although I couldn't 100% figure out how to use it properly. It has a hack to not do xauth if it does everything through ssh.
| |
| 11:39 | which is fine because neither case has -ac
| |
| 11:39 | it could use some cleanup but best to get a working version in sooner
| |
| 11:40 | <laga> johnny: 2.6.24
| |
| 11:40 | johnny: whaever is in ubuntu hardy :)
| |
| 11:58 | * warren still trying to remove that hack | |
| 12:01 | chupa has joined #ltsp | |
| 12:02 | chup has quit IRC | |
| 12:09 | <johnny> hmm.. i wonder if virtualbox additions will work with hardy this time..
| |
| 12:18 | <vagrantc> warren: i had another idea to test ... how does X handle if /etc/X11/xorg.conf is a symlink to a non-existant file? ... we could symlink it to /var/run/ltsp/xorg.conf and if we need to generate that file, do so, and if not, hopefully it will fall back to autodetection
| |
| 12:18 | <warren> vagrantc, uh....
| |
| 12:18 | * warren does a quick test | |
| 12:19 | <warren> vagrantc, "if we need to generate that file" is what case?
| |
| 12:19 | <vagrantc> warren: if the X resolution needs to be a specific size
| |
| 12:20 | i.e. X_MODE_0=1024x768 ... even if the monitor could do a higher resolution
| |
| 12:20 | but you have a class full of disparate hardware, and you want consistancy
| |
| 12:20 | <warren> i'm testing the broken symlink
| |
| 12:21 | <vagrantc> cool
| |
| 12:21 | <warren> although this is with X.org 1.4.99.something
| |
| 12:21 | it has all this futuristic stuff
| |
| 12:22 | <laga> flying skateboards? cool
| |
| 12:23 | <warren> laga, yes, but requires FAA approval and you need to go through TSA screening.
| |
| 12:23 | vagrantc, seems to properly autodetect when xorg.conf doesn't exist.
| |
| 12:23 | <laga> i've had it with these freedesktop.org guys!
| |
| 12:24 | <warren> vagrantc, although with configuration where you give it only the resolution, how does it figure out the driver and other option?
| |
| 12:24 | options
| |
| 12:24 | <vagrantc> warren: well, doesn't matter. we have a hook to allow you a configuration program of your choosing
| |
| 12:25 | <warren> Fedora doesn't HAVE a configuration program
| |
| 12:25 | <vagrantc> warren: the configure-x.sh script uses X itself for the autodetection, for example
| |
| 12:25 | warren: i KNOW.
| |
| 12:25 | <warren> =)
| |
| 12:25 | hm
| |
| 12:25 | xauth -f $XAUTHORITY list | to something
| |
| 12:25 | how can I guarantee that I get only one line?
| |
| 12:25 | <vagrantc> but is it such a bad thing to allow someone to have a configuration program if they want one?
| |
| 12:26 | warren: tail or head ...
| |
| 12:26 | <jammcq> head -1
| |
| 12:26 | <vagrantc> guaranteeing that it's the *right* line will be tricky though, if it matters
| |
| 12:26 | <jammcq> sed -n '1p'
| |
| 12:26 | <warren> hm
| |
| 12:26 | I'll just use head for now
| |
| 12:26 | you can see the code and figure out a better way
| |
| 12:27 | <vagrantc> i'm wondering if tail would be better ... assuming that the order is based on which is most recent
| |
| 12:27 | <warren> order is whatever you made it
| |
| 12:28 | vagrantc, so what is the standard location /etc/X11/xorg.conf should be pointing at?
| |
| 12:28 | I'm OK with whatever
| |
| 12:28 | * warren gets /whatever added to FHS | |
| 12:28 | <vagrantc> warren: well, i liked the /var/run/ltsp-xorg.conf ...
| |
| 12:29 | <warren> why not /var/run/xorg.conf?
| |
| 12:29 | <vagrantc> yeah, /whatever would be a good addition to the FHS ...
| |
| 12:29 | warren: well, we're ltsp, so just in case someone else wants to use that namespace ...
| |
| 12:30 | <warren> ok, whatever
| |
| 12:30 | <vagrantc> actually, since we're using multiple files under /var/run, we should probably start to use sub-dirs ... i.e. /var/run/ldm/ /var/run/ltsp /var/run/ltspfs ...
| |
| 12:31 | <warren> just pick something so we can be stuck with it forever
| |
| 12:31 | =)
| |
| 12:33 | <vagrantc> well, sure.
| |
| 12:34 | i've been trying to move towards something more permanent, yes.
| |
| 12:34 | but we keep changing what we're doing.
| |
| 12:35 | <warren> vagrantc, would you mind if I pushed the xauth changes before the screen script common stuff? it might conflict, but my additions are in one block in screen.d/ldm and should be easy to merge.
| |
| 12:37 | <vagrantc> warren: if it's all in ldm, that's fine. :)
| |
| 12:37 | warren: i want xauth as soon as possible :)
| |
| 12:38 | 90% of what changes in the ldm script is just removing everything
| |
| 12:39 | Pascal_1 has joined #ltsp | |
| 12:40 | <warren> nod
| |
| 12:44 | K_O-Gnom has quit IRC | |
| 12:47 | <warren> vagrantc, /var/run/ltsp/xorg.conf?
| |
| 12:47 | vagrantc, /var/run/ltsp/ltsp-xorg.conf seems redundant to me
| |
| 12:47 | <vagrantc> warren: works for me.
| |
| 12:49 | K_O-Gnom has joined #ltsp | |
| 12:50 | chup has joined #ltsp | |
| 12:50 | chupa has quit IRC | |
| 12:52 | * johnny builds ltsp-build-client on ubuntu | |
| 12:55 | <Pascal_1> hello !
| |
| 12:57 | * warren stuck on automake... | |
| 12:58 | K_O-Gnom has quit IRC | |
| 13:08 | K_O-Gnom has joined #ltsp | |
| 13:28 | mikkel has quit IRC | |
| 13:37 | subsume has joined #ltsp | |
| 13:42 | chup is now known as chupacabra | |
| 13:45 | <warren> vagrantc, I can't figure out how to use automake to install the rc.d file
| |
| 13:45 | <vagrantc> warren: look in ltspfs
| |
| 13:45 | warren: might be able to use something similar ...
| |
| 13:47 | <warren> I seem to be missing something obvious
| |
| 13:48 | ah, might have found it
| |
| 13:49 | had to modify configure.ac too
| |
| 13:51 | last test then I'll push it...
| |
| 13:54 | <vagrantc> the xauth changes? :)
| |
| 13:54 | Pascal_1 has quit IRC | |
| 13:57 | <warren> yes
| |
| 14:04 | <dberkholz> johnny, laga: tried the fuse unionfs?
| |
| 14:06 | <laga> no, i re-fused (someone kill me for that pun)
| |
| 14:06 | doesn't fuse have problems with applications like azureus?
| |
| 14:08 | <warren> vagrantc, pushed
| |
| 14:09 | <dberkholz> i don't really use bittorrent
| |
| 14:09 | wouldn't know
| |
| 14:16 | <laga> dberkholz: hum, i can't find any reference right now.. except for some breakage in feisty which has been fixed since
| |
| 14:20 | slipttees_ has joined #ltsp | |
| 14:25 | <vagrantc> warren: why on earth did you use sbalneav's /tmp/foople ?
| |
| 14:25 | <johnny> dberkholz, i was thinking about that
| |
| 14:25 | <warren> vagrantc, that's only debugging lines
| |
| 14:25 | vagrantc, it is REALLY difficult to know what's going on under the hood without that
| |
| 14:25 | <vagrantc> warren: sure, but ... /tmp/foople? :)
| |
| 14:25 | <warren> vagrantc, go ahead and fix it if you have a better idea =)
| |
| 14:26 | <vagrantc> warren: i also don't think you need to do the getltscfg call
| |
| 14:26 | warren: all that stuf is exported into the environment already
| |
| 14:27 | <warren> vagrantc, I just copied that from sbalneav's paste
| |
| 14:27 | vagrantc, go ahead and clean it up
| |
| 14:27 | <vagrantc> warren: thanks for working on it :)
| |
| 14:31 | otavio has quit IRC | |
| 14:32 | slipttees_ is now known as slipttee | |
| 14:33 | <johnny> dberkholz, genkernel4 has a nice setup in some ways, but very nasty in others.. and is seemingly unmaintained, and will not be accepted directly in genkernel upstream
| |
| 14:33 | genkernel3 dropped unionfs altogether, since it's only a kernel patch now, no loadable module
| |
| 14:33 | genkernel3's initrd scripts are kinda weak too
| |
| 14:34 | genkernel 3 folks are thinking about how to more modualarize genkernel for 3.5, but atm only agaffney and wolf32o1 are maintaining it
| |
| 14:35 | and they aren't very willing maintainers either..
| |
| 14:37 | <sbalneav> Afternoon all
| |
| 14:38 | <johnny> scottastic
| |
| 14:38 | welcome
| |
| 14:39 | dberkholz, i did find out my issue with dnsmasq and busybox not working well together if dnsmasq is on a different server than where the rootfs is solved in busybox 1.9x
| |
| 14:40 | <dberkholz> johnny: kernel patch and loadable module are not mutually exclusive...
| |
| 14:40 | <johnny> you can use -O root-path or whatever
| |
| 14:40 | i meant external module
| |
| 14:40 | <dberkholz> yes, i know, i was cc'd on that bug =)
| |
| 14:40 | <johnny> sorry
| |
| 14:40 | * vagrantc waves to sbalneav | |
| 14:40 | <sbalneav> Hey vagrantc!
| |
| 14:40 | <stgraber> hi sbalneav
| |
| 14:40 | <sbalneav> Hey stgraber!
| |
| 14:41 | <johnny> dberkholz, genkernel won't use busybox 1.9 until the mdadm patches get ported
| |
| 14:41 | do you happen to know why they aren't upstreamed?
| |
| 14:43 | <warren> sbalneav, thanks to your xauth paste I added the first working (although very ugly) version to ldm-trunk
| |
| 14:43 | sbalneav, including /tmp/foople, which scared vagrantc
| |
| 14:43 | <sbalneav> Ah, the /tmp/foople was just debug stuff
| |
| 14:43 | hose it.
| |
| 14:43 | <warren> sbalneav, well, I think it is still needed because while I got it working, it isn't quite behaving properly
| |
| 14:44 | <vagrantc> /var/log/ldm ?
| |
| 14:44 | <sbalneav> Anything I write that has "foo" "foople" "foobar" in it can usually be removed :)
| |
| 14:44 | <warren> sbalneav, for example, if I generate a xauth file before running X, ssh -Y doesn't work.
| |
| 14:44 | but there is no reason why that should be
| |
| 14:44 | it works fine with gdm or startx
| |
| 14:44 | slipttee is now known as slipttees_ | |
| 14:45 | <sbalneav> ah, so grab a hex key, generate the .Xauthority file first, then start X?
| |
| 14:45 | <warren> vagrantc, IMHO, just leave it as-is until we perfect the xauth usage, then remove it entirely.
| |
| 14:45 | sbalneav, that's how gdm and startx work
| |
| 14:45 | slipttees_ has quit IRC | |
| 14:45 | <jammcq> hey
| |
| 14:45 | <sbalneav> Hey jammcq!
| |
| 14:45 | * vagrantc misses ogra already | |
| 14:45 | <jammcq> there used to be a guy names sbalneav that hung around here
| |
| 14:46 | are you any relation?
| |
| 14:46 | <sbalneav> He's here right now! :)
| |
| 14:46 | Where's ogra?
| |
| 14:46 | <jammcq> Scotty, is it really you ?????????????????????????????????????????
| |
| 14:46 | <sbalneav> Yep!
| |
| 14:46 | <jammcq> someone pissed him off
| |
| 14:46 | <vagrantc> sbalneav: the "so long" thread on ltsp-developer
| |
| 14:46 | <sbalneav> Who did that?
| |
| 14:46 | What?!
| |
| 14:46 | I haven't read for a few days.
| |
| 14:46 | <vagrantc> we get back sbalneav only to loose ogra. what sort of cruel world is this?
| |
| 14:47 | <jammcq> we'll get ogra back
| |
| 14:47 | * vagrantc hopes it is only temporary | |
| 14:47 | <johnny> yes.. bring on the ogra..
| |
| 14:47 | swell fella that one..
| |
| 14:47 | <sbalneav> I'm just digging into seeing that it would take to get sshd to run a graphical password changer
| |
| 14:47 | <johnny> unlike this scott character :)
| |
| 14:48 | <warren> sbalneav, so ldm-trunk currently only generates the xauth only if you are using LDM_DIRECTX
| |
| 14:48 | sbalneav, that way both with and without work, and neither need -ac
| |
| 14:48 | I'm posting to list soon the further TODO's.
| |
| 14:51 | <dberkholz> johnny: no, i don't know why
| |
| 14:54 | <johnny> uggh.. if i figure this thing out.. i might have it booting
| |
| 14:56 | genkernel4 has really nice features.. but all the custom snaps of projects aint helping it be acceptable to anybody else, and is very prone to breakage
| |
| 14:58 | <vagrantc> johnny: genkernel handles what?
| |
| 14:59 | <johnny> kernel and initramfs
| |
| 15:00 | a few developer started working on genkernel4, but then the main one (if i understand correctly) retired, and now the project is external
| |
| 15:00 | and not touched since 11/17/07
| |
| 15:01 | chupacabra has quit IRC | |
| 15:01 | chupacabra has joined #ltsp | |
| 15:12 | Administrateur has joined #ltsp | |
| 15:14 | <warren> vagrantc, wait a sec...
| |
| 15:14 | vagrantc, I'm not against a symlink to /var/run/ltsp/xorg.conf per se
| |
| 15:14 | vagrantc, but what does this gain us over running X with -config?
| |
| 15:15 | <vagrantc> warren: if they're not using an LTSP script to start it, it's where X expects it to be
| |
| 15:15 | <warren> ah.
| |
| 15:15 | ok
| |
| 15:15 | <vagrantc> like running all applictions locally, GDM won't require any hacking.
| |
| 15:15 | RyanRyan52 has quit IRC | |
| 15:16 | <vagrantc> warren: i'm not 100% sure we should bother with it ...
| |
| 15:16 | warren: it was just an idea i wanted to throw out there
| |
| 15:16 | <warren> if you are running all applications locally, I wouldn't bother
| |
| 15:16 | each distro handles this case in its own way
| |
| 15:22 | RyanRyan52 has joined #ltsp | |
| 15:24 | <johnny> dberkholz, genkernel4 moved
| |
| 15:24 | they might rename it
| |
| 15:28 | <vagrantc> warren: ok, let's keep doing it the way we have been, and but we know that's an option for cases where it might make sense.
| |
| 15:28 | <dberkholz> johnny: where's the code now?
| |
| 15:28 | <vagrantc> warren: er, way we have been meaning passing -config in the screen scripts
| |
| 15:28 | <dberkholz> i know they picked up a domain, but nothing on the web yet
| |
| 15:28 | <vagrantc> that being kind of a new idea anyways
| |
| 15:28 | <johnny> http://svn.berlios.de/wsvn/genkernel/
| |
| 15:30 | praveer_cool has quit IRC | |
| 15:31 | Administrateur has quit IRC | |
| 15:32 | <dberkholz> good to know
| |
| 15:33 | <johnny> i'm going to give it a shot
| |
| 15:36 | Q-FUNK has quit IRC | |
| 15:37 | praveer_cool has joined #ltsp | |
| 15:48 | vagrantc has quit IRC | |
| 16:06 | Q-FUNK has joined #ltsp | |
| 16:09 | jhp has quit IRC | |
| 16:36 | johnny_ has joined #ltsp | |
| 16:40 | mistik1 has quit IRC | |
| 16:52 | mistik1 has joined #ltsp | |
| 16:53 | johnny_ has quit IRC | |
| 16:54 | Q-FUNK has quit IRC | |
| 16:55 | johnny_ has joined #ltsp | |
| 16:55 | chupacabra has quit IRC | |
| 16:56 | chupacabra has joined #ltsp | |
| 17:17 | Q-FUNK has joined #ltsp | |
| 17:46 | joebake1 has joined #ltsp | |
| 17:46 | joebake1 has left #ltsp | |
| 17:47 | joebake1 has joined #ltsp | |
| 17:51 | chup has joined #ltsp | |
| 17:53 | chupacabra has quit IRC | |
| 17:56 | K_O-Gnom has quit IRC | |
| 17:56 | Q-FUNK has quit IRC | |
| 18:12 | <slashdot1x> yeah it's working now
| |
| 18:13 | putting modules blacklist under /etc/modprobe.d
| |
| 18:13 | and later on "update-initramfs -u" does the work
| |
| 18:23 | vagrantc has joined #ltsp | |
| 19:07 | <vagrantc> warren: pushed the common-x-screen script stuff
| |
| 19:07 | working on ldm now
| |
| 19:08 | mccann has joined #ltsp | |
| 19:20 | <vagrantc> i should have put this into a real branch rather than just shelving the changes...
| |
| 19:44 | cliebow has joined #ltsp | |
| 19:59 | J45p3r has joined #ltsp | |
| 20:12 | johnny_ has quit IRC | |
| 20:21 | MacIver has quit IRC | |
| 20:24 | cliebow has quit IRC | |
| 20:24 | MacIver has joined #ltsp | |
| 20:33 | slashdot1x has left #ltsp | |
| 20:36 | J45p3r has quit IRC | |
| 20:36 | J45p3r has joined #ltsp | |
| 20:37 | talntid has joined #ltsp | |
| 20:37 | <talntid> Hi all :)
| |
| 20:37 | I: Base system installed successfully.
| |
| 20:37 | error: LTSP client installation ended abnormally
| |
| 20:37 | Not sure where to go from here.... :)
| |
| 20:39 | <vagrantc> talntid: linux distro and release?
| |
| 20:39 | J45p3r has quit IRC | |
| 20:43 | <talntid> Ubuntu 7.10
| |
| 20:43 | running a Xen kernel.. so LTSP is on a virtual machine..
| |
| 20:43 | LTSP 5.0.39
| |
| 20:45 | jblack has joined #ltsp | |
| 20:45 | <talntid> I am using WyseOS thin clients, and the server is AMD64.. should I be using an AMD64 image or i386?
| |
| 20:48 | nobody? :)
| |
| 20:50 | <jammcq> well
| |
| 20:50 | that's a tricky question
| |
| 20:51 | I'd prolly load the i386 version on the server, because the users will be running desktop sessions on the server. for the most part, 64-bit works ok, until you get into running flash or acroread or other non-free things
| |
| 20:52 | <talntid> so the AMD64 would work on the thin clients, even though the thin client hardware is not 64bit?
| |
| 20:52 | <jammcq> if you do install the 64-bit os, you'll want to still install the 32-bit version of LTSP, cuz that's what the thin clients will ned
| |
| 20:52 | <talntid> right.
| |
| 20:52 | i misunderstood your first question
| |
| 20:52 | i mean answer.
| |
| 20:52 | <jammcq> so, it's entirely possible to have 64-bit on the server, and 32-bit for the clients
| |
| 20:52 | <jblack> Ok. So the thing clients are doing more than just rendering an x display. i.e. the processing happens on the thin clients
| |
| 20:53 | <jammcq> well, not really
| |
| 20:53 | actually, the thin clients ARE just running enough code to boot up and run an X server
| |
| 20:53 | all apps are running on the server
| |
| 20:53 | <jblack> Ok, so when we do build-image, is that just the code to get the xserver going and such?
| |
| 20:54 | <jammcq> the thing you need to figure out is, are the apps YOU want to run, available in 64-bit ?
| |
| 20:54 | yes, the build-image sets up the image for the thin client. when you run that, you'll want the 32-bit
| |
| 20:54 | <talntid> the apps I want to run are Firefox. and OpenOffice. :)
| |
| 20:54 | <jblack> Most of them should be, yes, excepting possibly flash.
| |
| 20:54 | <jammcq> ok, firefox and Openoffice are fine in 64-bit. but, will the users be browsing to sites that have flash animation?
| |
| 20:55 | <talntid> sometimes.
| |
| 20:55 | possibly.
| |
| 20:55 | <jammcq> and will they want acroread for viewing PDF's ?
| |
| 20:55 | <talntid> yes
| |
| 20:55 | or anoyher PDF viewer
| |
| 20:55 | another
| |
| 20:55 | <jammcq> well, you could use the gnome pdf viewer, but personally, I prefer acroread
| |
| 20:55 | <talntid> and it's not been ported to AMD64?
| |
| 20:56 | <jammcq> those are the types of problems you could run into, if you go 64-bit
| |
| 20:56 | <talntid> I see.
| |
| 20:56 | Where is a log file to see why the ltsp-build-client --arch i386 is ending abnormally?
| |
| 20:56 | <jammcq> acroread isn't 64-bit, as far as I know. I believe there is ways to run the 32-bit versions of those things on 64-bit linux
| |
| 20:56 | ummm, I dunno
| |
| 20:56 | <jblack> It may be possible to run a xen domu in 32 bit, even though the rest of the system is 64
| |
| 20:56 | <jammcq> you need a ltsp-5 guy to answer that. not sure if there's any around right now
| |
| 20:57 | <talntid> ok :)
| |
| 20:59 | petre has joined #ltsp | |
| 20:59 | <petre> evening all
| |
| 21:02 | <talntid> hi :)
| |
| 21:02 | are you a LTSP-5 guy? :)
| |
| 21:03 | <petre> no, I'm not really up to speed on 5, more familiar with 4.2
| |
| 21:03 | <talntid> d'h :)
| |
| 21:03 | <petre> been away from it for the past several months, but hoping to get back into it
| |
| 21:04 | <talntid> so, if on 4.2...
| |
| 21:04 | when building it says...
| |
| 21:04 | I: Base system installed successfully.
| |
| 21:04 | error: LTSP client installation ended abnormally
| |
| 21:04 | <vagrantc> there is no ltsp-build-client is not part of 4.2
| |
| 21:04 | <talntid> oh. ;)
| |
| 21:04 | thanks.
| |
| 21:04 | * talntid sits in the corner. | |
| 21:05 | <vagrantc> don't know off the top of my head why --arch wouldn't work, but it's definitely not well tested.
| |
| 21:05 | <petre> vagrantc, you still working for canonical?
| |
| 21:05 | <vagrantc> petre: never have.
| |
| 21:05 | <johnny> you mean ogra probably..
| |
| 21:05 | yes he is
| |
| 21:06 | <vagrantc> petre: came close, once upon a time :)
| |
| 21:06 | <petre> I know about ogra; I thought you were just about to work for them, oh, a year or so ago; my mistake
| |
| 21:07 | my recollection was that you were in Spain at the time or thereabouts
| |
| 21:07 | <vagrantc> petre: you're recollections are right, but in the end it didn't pan out.
| |
| 21:08 | <petre> so who's paying your bills these days?
| |
| 21:08 | contract stuff?
| |
| 21:08 | <vagrantc> not nearly as many people as i would like :)
| |
| 21:09 | a little subbing work at freegeek (http://freegeek.org)
| |
| 21:09 | <johnny> what do you do vagrantc ?
| |
| 21:09 | <petre> I know how that goes
| |
| 21:10 | <vagrantc> johnny: i mostly sit around working on ltsp all the time :)
| |
| 21:10 | <petre> my current employer is a linux shop, a good thing, but no LTSP *sigh*
| |
| 21:10 | <johnny> freegeek pays you to do that?
| |
| 21:10 | <vagrantc> johnny: not so much ltsp, but i do sysadmin and various floor shifts for freegeek.
| |
| 21:11 | <johnny> well.. what are you skills other than ltsp
| |
| 21:11 | <vagrantc> i can sometimes sneak in a little ltsp work, as freegeek uses ltsp.
| |
| 21:11 | <johnny> in case i run itno things
| |
| 21:11 | i've had opportunities for jobs, just nobody to trust them with
| |
| 21:12 | <vagrantc> johnny: probably best outside of #ltsp :)
| |
| 21:13 | <johnny> if you think so..
| |
| 21:13 | sure.. the jobs themselves
| |
| 21:13 | i just want to know what you can do
| |
| 21:13 | if you wanna pm it.. go ahead
| |
| 21:28 | praveer_cool has quit IRC | |
| 21:55 | petre has quit IRC | |
| 22:09 | swarm2 has joined #ltsp | |
| 22:10 | swarm2 has quit IRC | |
| 22:17 | chupa has joined #ltsp | |
| 22:19 | chup has quit IRC | |
| 22:20 | talntid has quit IRC | |
| 22:25 | RyanRyan52 has quit IRC | |
| 22:28 | RyanRyan52 has joined #ltsp | |
| 22:30 | RyanRyan52 has quit IRC | |
| 22:33 | RyanRyan52 has joined #ltsp | |
| 22:37 | subsume has quit IRC | |
| 22:52 | loather-work has quit IRC | |
| 23:04 | chup has joined #ltsp | |
| 23:04 | chupa has quit IRC | |
| 23:11 | vagrantc has quit IRC | |
| 23:20 | basanta has joined #ltsp | |
| 23:31 | daya has joined #ltsp | |
| 23:54 | chupa has joined #ltsp | |
| 23:56 | chup has quit IRC | |